Source: BAM
The construction industry is one of the largest producers of waste in Europe and is responsible for around ten percent of CO2 emissions. In addition, there is a large amount of construction and demolition waste, which accounts for 25-30 percent of all waste in Europe. The research group led by Prof. Dr. Sabine Kruschwitz is investigating how materials can be recycled and processes optimized to support a circular economy wherever possible.
Team
Approximately 10 - 15 employees, including postdocs, PhD students, bachelor, and master students, but also guest scientists, focus on innovative topics in non-destructive material characterization and material informatics of building materials and thus complement the scientific portfolio of the department Non-destructive Testing.
Topics
Specifically, the research group is advancing the following topics:
I) Material characterization to support
- The development of sustainable building materials by the identification of new approaches for the early evaluation of durability properties
- The development of efficient tools for characterizing the early hydration behavior/rheology of CO2-reduced cement mixtures
- The development of (multi-method) characterization solutions for the circular economy in the construction industry
II) Materials informatics to accelerate
- The discovery of environmentally friendly building material compositions through a data-driven approach
- The development of AI software tools for materials science applications
- The creation, integration, enrichment, and use of semantic material data structures
Methods
The following methods of material characterization are used:
Low-field 1H nuclear magnetic resonance (NMR) at BAM
NMR relaxometry is used for material characterization and for the determination of durability-describing parameters such as porosity, permeability and pore size distribution. In view of the increasing demand for resource-efficient and CO2-reduced materials, NMR enables the monitoring of hydration processes and thus supports the evaluation and optimization of newly developed cement mixtures. Available low-field NMR instruments at BAM are the NMR-MOUSE1 from Magritek Ltd. and the MR Core Analyzing Tomograph2 from Pure Devices GmbH.
[1] https://magritek.com/products/nmr-mouse/ (09.07.2020)
[2] https://www.pure-devices.com/images/Flyer/Flyer_MR-CAT.pdf (07/09/2020)
Spectral Induced Polarization (SIP)
The spectral induced polarization (SIP) method is a well-established geophysical method that our group has been using as a non-destructive testing method in civil engineering for more than 15 years. Our work mainly focuses on understanding fundamental relationships (e.g., texture-related control of the characteristic frequency dependence of materials) through laboratory measurements under well-controlled conditions.
Other methods in cooperation with other departments:
- Investigation by use of µ-CT together with the departments of Radiological Methods and Micro-Destructive Testing (ZFP).
- Application of Laser Induced Breakdown Spectroscopy (LIBS) together with the Department of Non-Destructive Testing Methods for Civil Engineering
- Application of Heat Flow Calorimetry together with the Department of Building Materials Technology
Application examples and current research
We are currently working on the following research questions:
Evaluation of recycled building materials
The average lifetime of a building in Europe is just under 40 years, after which it is demolished. The reason is often so-called functional obsolescence: If a building or even only a part of it is to be reused or repurposed, the necessary information on the original construction method is missing for approval. This practice results in a large amount of construction and demolition waste, which accounts for 25-30 percent of all waste in Europe. The major European project "Reincarnate", funded by the EU's Horizon Europe research and innovation program, aims to change this unsustainable use of resources.
Contact: Sabine Kruschwitz
Read the press release from 2022 here (in German).
Hydration behavior of cement-reduced mixes
Understanding the specific hydration process of new cement mixtures is the key to successful material development. During this process, the research is also exploring new methods for evaluating and classifying building material properties. In our work, NMR is used to investigate the hydration process of alternative, climate-friendly binders and to compare them with those of known compositions.
Further information about the project and contact persons can be found here.
Transport processes of moisture and salt ions in porous building materials analyzed by NMR and LIBS
Measuring the transport properties of cement-based porous building materials is important to understand their mechanical properties and durability. In our work, we use a combination of NMR and laser-induced breakdown spectroscopy (LIBS) to study the transport properties of different pore systems and brine concentrations and attempt to determine material-specific transport coefficients based on our observations.
More information about the project and contact persons can be found here.
Supporting the circular economy: Sorting construction and demolition waste in "LIBS-ConSort”
Closed material cycles and sorted material fractions are necessary to achieve high recovery and recycling rates in the construction industry. In the recycling of construction and demolition waste, simple but proven techniques have been preferred in the past to process large quantities of construction waste in a short time. This is in contrast to the increasing complexity of composites and structures in the mineral building materials industry. Automated, sensor-based sorting of construction materials could complement or replace the current manual, and therefore dangerous and costly, practice to improve processing speed, recycling rates, sorting quality and health conditions.
Further information on the project and contact persons can be found here.
Fundamental research: Interpretation of NMR signals in partially saturated media
The NMR relaxation time of hydrogen protons in porous materials depends on numerous factors. In addition to molecular bonding, the relaxation behavior is also influenced by the environment, for example, the pore geometry and the surrounding mineral composition.
To also be able to assign the NMR signals of partially saturated systems to the pore sizes, we are currently developing an approach at BAM which is based on adsorption and includes the calculation of the adsorbed water layer thickness.
Further information on the project and contact persons can be found here.
Multi-method analysis of pore spaces in sandstones
To estimate the moisture and pollutant ion transport properties of porous building materials, their pore size distributions are often used. These can be determined by various methods such as mercury intrusion porosimetry (MIP) or imaging techniques such as micro-computed tomography (µ-CT), which, however, are not suitable for in-situ use in the field. Nuclear magnetic resonance (NMR) offers this possibility.
In our research, we use a systematic multimethod approach based on a combination of MIP, NMR, µ-CT, and Spectral Induced Polarization (SIP), which has been applied mainly to sandstones yet. We have been able to show for some of the samples studied that only the joint evaluation of the four methods provides a complete understanding of the characteristic lengths. This approach will now be generalized and transferred step by step from samples with first simpler and then more complex pore systems.
Contact: Sabine Kruschwitz
Data Science: AI App "SLAMD"
SLAMD is an open-source app that takes a revolutionary approach to optimizing the process of material design - particularly in the development of sustainable building materials. Due to the increasing complexity of building material formulations, it is becoming more and more difficult to find the ideal formulation through conventional trial and error methods. However, using global optimization and inverse design techniques, all possible formulations can be searched to find new and improved approaches. The app is based on sequential learning, a type of artificial intelligence that learns from empirical observations and simulations, and is thus able to go beyond already known materials and find new, more sustainable solutions with less data. SLAMD maps all the necessary steps of the workflow, thereby lowering application barriers.
More information about the project and contact persons can be found here.
Interview with Christoph Völker
Video "Sustainable Building with AI" at Berlin Science Week 2022.
Data Science: Ontologies
Measurement data is an important basis for scientific analysis in the field of non-destructive testing (NDT) in the construction industry, but until now there has been no standardized representation of this data. Therefore, an analysis of data sets across different test objects or test types requires a high manual effort. We investigate how intelligent classification schemes, so-called ontologies, can be used to make complex data efficiently searchable. While data searchability depends on the definition of search terms, the semantic rules defined by the ontology, which take into account relationships within a domain, enable even more comprehensive search across disciplines.
We are therefore systematically developing knowledge on how intrinsic features of NDT measurements can be represented in an application-oriented way using semantic technologies. From this, we develop reference workflows and ontology design patterns that we test for effectiveness in research practice and make available for future work in this area.
Contact: Benjamin Moreno Torres